Production and perception of L2 English orthographic and phonological representations by L1 Tera/Hausa speakers :an experimental study

Abstract

Oral English is an integral part of the English language syllabus in Nigeria's secondary schools. However, pronunciation is problematic for students due to factors which include the complexity of grapheme-phoneme correspondences in English, the influence of the students' first language (L1), Tera/Hausa and the method of teaching. Research in second language (L2) acquisition of phonology and on the role of orthographic input has shown that learners’ phonological development can be affected as a result of L2 orthographic input (e.g. Young-Scholten 2002; Rafat 2011 & 2016; Bassetti, Escudero and Hayes-Harb 2015; Bassetti and Atkinson 2015; Young-Scholten and Langer 2015). To this effect, the present study is based on the idea that it is possible to address the difficulties that teaching L2 English pronunciation creates for L1 Tera/Hausa (Chadic languages) speakers. This involved providing L2 orthographic input to see whether it influenced learners’ underlying representations and in turn their oral production. The research involved an intervention study which was aimed at experimentally examining Tera/Hausa speaking secondary school students' production and perception of English orthography and phonological representations with respect to consonant clusters such as clock, straw, and desk, digraphs in clusters such as bench, fridge, and syringe, silent singletons such as knife, signboard, and whistle and digraph singletons such as phone, duck, and ring. The study was conducted with 73 Tera/Hausa speaking secondary school students in Gombe state, Nigeria in pre-testing and post-testing in four sub-tests consisting of two production tasks (picture-naming and reading aloud) and two perception tasks (epenthesis and dictation). A proficiency test was conducted prior to the pre-test which resulted in participants being identified as belonging to three proficiency levels. The learners were randomly (not based on their proficiency) divided into three experimental condition groups and taught eight lessons in 20-minute sessions over four school weeks. Three methods were used for the instruction: listening + orthography group where the group were taught while listening to native speaker recordings of the lessons on an audio player while seeing the written forms; listening-only group, who were taught by only listening to the recordings of the lessons without any written form; and traditional teaching who were taught by a non-native speaking English teacher using the teaching methods normally used in Nigeria to teach English. The hypotheses for the study were based on the idea that although Tera and English both use the Roman alphabet they have their own orthographies, grapheme-phoneme correspondences differ and this will affect Tera/Hausa speakers’ L2 phonology. As a result, without intervention iv at pre-test, it was predicted that the learners will not correctly produce and perceive L2 English consonant clusters, digraphs in clusters, silent singletons and digraph singletons due to problems with the L2 syllable structure. However, with intervention among three experimental condition groups, there will be significant improvement by the group that received explicit phonological and orthographic input than the other groups which did not receive explicit instruction. Qualitative analysis revealed a greater reduction at post-test in error rate by the listening + orthography group on all the error categories on the picture-naming task, reading aloud task and dictation task, compared to the traditional teaching method group and the listening-only group. There was a scattered error reduction rate by the three different proficiency levels. Similarly, in the quantitative analysis, the listening + orthography group yielded significantly greater improvement on the dictation task, picture-naming task and reading aloud task (p ≤ 0.05) compared to the traditional teaching method group and the listening-only group. Although, the traditional teaching method group yielded better improvement on the epenthesis task, the difference between their mean scores with that of the listening + orthography group did not differ significantly (0.22 points). Proficiency level, however, did not significantly influence performance on any of the tasks. The study highlights the effect of orthographic input on Tera/Hausa learners’ production perception in the acquisition of English and uniquely serve as the first phonological acquisition study with African data. The findings of this study allow us to make recommendations for the best and most effective ways of teaching oral English in Nigeria and in secondary schools elsewhere.